Odds ratio and effect sizes based on the result of Logistic regression model:
Note: You will plug in two values from the logistic regression result: the logit coefficient from the intercept and the logit coefficient for the treatment variable (program impact coefficient).
The odds ratio, however, does not depend on the intercept as it is simply the exponential of the logit coefficient. In Excel function:
ODDS_RATIO= exp( < here, plug in the logit coefficient > )
Adjusted percentages derived for the treatment and control groups, however, depend on both the intercept and program impact coefficient. The percentages will help readers to understand the meaning of results. Notice, however, the meaning of the intercept depends on what the reference category (aka omitted category) is.
T-test for binary variables (comparison of proportions)